The Importance of Stimulating Client Online Reviews Jacques Bulchand-Gidumal a, Santiago Melián-González a, and Beatriz González López-Valcárcel a a Faculty of Economics, Business and Tourism, Las Palmas de Gran Canaria University, Canary Islands, Spain {jbulchand, smelian, bvalcarcel@dede.ulpgc.es} Abstract The use of websites to solicit and share opinions on products consumed is now a standard practice. There is reason to believe that dissatisfied customers are those most likely to recount and evaluate their experiences on these specialized sites, and it is therefore in the interest of businesses to encourage satisfied clients to do so as well. With a sample of 16680 hotels that belong to 249 tourist areas taken from TripAdvisor, this study shows that the fewer the reviews a hotel receives, the more negative these reviews will be. Additionally, it is proven that there is a chronological sequence: negative reviews arrive first. Thus, we recommend that those in charge of hotels and those promoting tourist destinations encourage the participation of hotel clients in the websites in which the establishments are commercialized and/or reviewed in order to obtain on the whole more favourable ratings. Keywords: online reviews; TripAdvisor; social media; hotels; customer management; tourism. 1 Introduction The tourism sector has evolved significantly with the spread of the Internet as has happened in every other sector and in social relations (Buhalis & Licata, 2002). In this new management cycle there is an increase in the number and quality of choices available to consumers, and in this context numerous websites have emerged to display client reviews and opinions of their experience. The traditional word-ofmouth mechanism for influence (WOM) has its equivalence in the information age, e- word-of-mouth (ewom) (Litvin et al., 2008). This study is based on the hypothesis that dissatisfaction and negative experiences generate more client WOM activity than satisfaction and positive experiences; so our goal is to see whether the quantity of client reviews affects the score of the reviews an establishment receives. 2 Online Reviews For the tourist sector there are several websites exclusively consisting of client reviews of services received, including lodging, transportation, and dining. Of them, TripAdvisor (www.tripadvisor.com) is one of the ones that stand out. The aim of these businesses is to provide independent reviews and opinions by users of tourist services to prospective tourists. It represents the largest travel community in the world, with more than thirty-five million visitors per month (coscore Media Metrix,
May 2010), and contains more than thirty-five million reviews and opinions on more than one million hotels, restaurants and attractions (TripAdvisor.com, 2010). Recent studies point to the influence of online reviews in consumer decisions. The Ripple-6 and e-tailing Group (www.ripple6.com, 2009) found that 83% of online buyers were interested in sharing their buying experience with people they knew, and that 41% would be willing to do so in online communities. 74% of these online buyers agreed that the exchange of information served like prior opinions to influence the choice of purchase. In the hotel sector, the results of the Second Fiturnews Survey showed the online opinion of other consumers was the most important variable in lodging choice for business trips, with 45.9% of those surveyed indicating that online reviews were the single factor that counted most for them (Fiturnews, 2008). 2.1 ewom WOM consists of client communications in regard to consumption with members of their social and professional network (Anderson, 1998). Its online equivalent, ewom, can be defined as "all informal communications directed at consumers through Internet-based technology related to the usage or characteristics of particular goods and services, or their sellers." (Litvin et al., 2008: 461). One of the determinants of WOM, and hence of ewom, is the level of satisfaction and the kind of emotions experienced in consumption (Zeelenberg and Pieters, 2004). Hence ewom is one of the behavioural responses to the dissatisfaction that predominates in the literature related to client satisfaction (Oliver, 1997). In this sense, WOM, including ewom, may be a behaviour more frequent among dissatisfied clients as opposed to those who are satisfied. Why this is so may be explained by the usefulness of WOM for those who engage in it. In this sense, following Zeelenberg and Pieters (2004), WOM can be a way to discharge dissatisfaction with a service or product (for example, disappointment) and obtain understanding and solidarity from others. These same authors point to the warning service that WOM provides in the case of dissatisfaction (Zeelenberg and Pieters, 2004: 449): "Customers may want to alert others about bad service providers." Should negative ewom indeed be more prevalent when the number of reviewer clients is low, it would make sense to stimulate ewom in all clients and thus obtain a higher rate of reviewing. Due to the current penetration of Internet in the society and the huge amount of consumers who use it, ewom is not depend on the skills needed to participate in a website, but on the benefits people feel they obtain by doing so (Parra-López et al., 2010). 3 Hypotheses and Methods The two hypotheses for this study are based on the proposition that negative ewom predominates among hotels with few reviews because out of all clients it is those dissatisfied who tend to WOM behaviour. So if the satisfied ones do not write reviews, the result is fewer and more negative reviews.
Hypothesis 1. The average rating that hotels receive on the Internet depends on the number of clients that review it and is lower for hotels with fewer reviews and higher for those with more reviews. Given that, should the above hypothesis hold, the doubt may arise whether in fact hotels with fewer reviews are indeed worse than those with more, we posit the following hypothesis. Hypothesis 2. Among the reviews that hotels receive, the first ones are more negative than the ones that are added later on, and this is because the first reviews tend more to be negative ewom than later one. Testing these hypotheses requires counting with a large sample so the results will be representative and robust. For this purpose we turned to TripAdvisor, mentioned at the opening of this study, the website we believe to be at present the most important for online reviews of lodgings. In March, 2010, we directly downloaded from TripAdvisor using automatized procedures all the online reviews of hotels in the 200 best-rated tourist destinations of Europe. Destinations are rated considering the hotels that belong to them, but also other factors such as the cultural offer, shopping, dining, nature, etc. The reviews in TripAdvisor are summed up in an overall rating in five possible categories: excellent (5), very good (4), average (3), poor (2) and terrible (1). The 200 destinations (e.g. Gran Canaria) were broken down into the different tourist zones they included, resulting in a total of 605 zones (e.g. Maspalomas, Playa del Inglés, and so on). Finally, in order to obtain a sample appropriate for the purpose of this study, as a criterion for inclusion of the hotels in the sample we chose only those with more than 10 reviews, and only zones with more than ten hotels, which resulted in a total of 249 tourist zones and 16680 hotels. In order to test hypothesis 2, we made a random sample of 423 hotels among the 2593 hotels that had between 101 and 200 reviews, since this number of reviews would allow us to compare the chronological evolution of reviews. Data on the sample is summarized in Table 1. Table 1. Information on the sample Source TripAdvisor Date of data collection March 2010 Data collection Automated using MS Excel Hotels with more than 10 reviews 16680 Hotels with 101 to 200 reviews 2593 Tourist zones with more than 10 hotels 249 Tourist destinations 200 Average number of reviews per hotel 74.8 % of very negative reviews (value 1) 4.9% % of very positive reviews (value 5) 13.9% To test the first hypothesis, about the hotel reviews, we grouped hotels in intervals according to the number of reviews received (i.e., those with from 11 to 20 reviews, from 21 to 30, etc.). These groups were defined so each would have a large amount of hotels and the size of each group would be of similar magnitude. Normality in each
group was tested. For the second hypothesis about the contrast between earlier and later hotel ratings we ordered the hotel reviews by date. We used ANOVA variance analysis software and Bonferroni's multiple comparison test to evaluate Hypothesis 1, since we wanted to compare the means of several groups to find if they were statistically different. A t-contrast of equal averages for independent samples for Hypotheses 2 was used. Calculations were made with STATA v11, with the significance level of 5%. 4 Results Table 2 shows a comparative analysis of the scores received by the hotels according to the number of reviews received. The average score improves with the number of reviews received, from 3.588 for the hotels with from 11 to 20 reviews, to 3.977 for the hotels with more than 100 reviews. The last three columns show that the percentage of the very dissatisfied (the score of 1 on a scale of 1 to 5) declines as the number of reviews increases, from 7.1% down to 3.4%. Similarly, as the number of reviews received increases, the percentage of those dissatisfied (scores 1 and 2 out of 5) drops from 17.6% to 10%, and conversely the percentage of score of excellent (5 out of 5) increases from 21.4% to 35.2%. By Bonferroni's test there are significant differences between all the groups, except between those with 41-50 and 51-100 reviews. Table 2. Ratings of hotels by the number of reviews Number of reviews Hotels Average % Very dissatisfied % Dissatisfied % Very satisfied 11-20 3461 3.588 7.14 17.7 21.4 21-30 2351 3.643 6.66 16.7 23.8 31-40 1699 3.711 5.55 15.4 25.0 41-50 1333 3.804 4.65 14.3 27.7 51-100 3853 3.794 4.76 13.5 27.4 101+ 3983 3.978 3.37 10.1 35.2 Total 16680 3.778 48.8 13.9 27.2 Table 3 shows comparisons between the average rating of the first 20 reviews received by the 423 hotels analyzed, and the average ratings when the number of reviews accumulate. Except for the comparison of the mean of the first 20 and the first 30 reviews (p=0.065), the rest of the comparisons demonstrate with statistical significance that as hotels receive more reviews, their average ratings improve, and the standard deviation declines, which means there is a greater agreement in ratings. Table 3. Evolution of the average ratings of reviews as the number of reviews increases Reviews considered Mean Standard deviation Comparisons First 20 3.743 0.722 20-30, p=0.065 First 30 3.756 0.665 30-40, p=0.007 First 40 3.772 0.628 40-50, p=0.041 First 50 3.780 0.608 50-100, p= 0.000 First 100 3.817 0.550 20-100, p=0.000
5 Conclusions This study confirms that as the number of reviews of a hotel increase, the ratings in these reviews are more positive. It was already known that dissatisfied customers are more motivated to evaluate. What this study proves is that good reviews also arrive, it is just they arrive later. Hence negative ewom is more prevalent when fewer clients evaluate the hotels. Thus both the directors of hotels as well as those administering and promoting tourist zones and destinations should make an effort to obtain a substantial number of reviews from their clients. Hotels should do because they will thus dilute the weight of negative reviews from disgruntled customers. Those responsible for tourist zones should also get involved, as the ratings for the destination will surely depend on the number of ratings of the hotels it includes. They should give some kind of recognition to the hotels that obtain a high number of reviews and facilitate the means for obtaining reviews, should these be beyond the capability of the hotels. Options could include free WIFI zones in the zones and airports, since, as the availability of technology affects the use of social media (Parra-López et al., 2010). Hotels must find ways to obtain online reviews without recurring to complex mechanisms for compensation. Different studies (Beenen et al., 2004; Parra-López et al., 2010) provide suggestions such as: taking advantage at the time of check-out to indicate to the client the importance of their particular opinion on given websites and have computers handy; indicating the importance of reviewing in a printed message on the invoice or receipt as a reminder should the client prefer to do it at another moment; sending a link to the review website in a email that emphasizes the importance of client participation within a given timeframe; or providing free WiFi so the clients can fill out a review with their own computers. All are simple procedures that can stimulate online reviews and thus compensate for the negative bias of ewom that is spontaneous. References Anderson, E.W. (1998). Customer satisfaction and word of mouth. Journal of Services Research 1: 5-17. Beenen, G., Ling, K., Wang, X., Chang, K., Frankowski, D., Resnick, P. & Kraut, R. (2004). Using social psychology to motivate contributions to online communities. Proceedings of ACM CSCW 2004 Conference on Computer Supported Cooperative Work, Chicago, IL, 212-221. Fiturnews (2008). http://www.mirahoteles.com/fitur2008/ Litvin, S.W., Goldsmith, R.E. & Pan, B. (2008). Electronic word-of-mouth in hospitality and tourism management. Tourism Management 29(3): 458-468. Parra-López, E., Bulchand-Gidumal, J., Gutiérrez-Taño, D. & Díaz-Armas, R. (2010). Intentions to use social media in organizing and taking vacation trips. Computers in Human Behavior doi:10.1016/j.chb.2010.05.022 Oliver, R.L. (1997). Satisfaction: a behavioral perspective on the consumer. New York: McGraw-Hill. Ripple6.com, (2009). http://www.ripple6.com/retail-shopping-communities-attract-shoppersinfluence-purchasing-and-retain-consumers.html